Transforming provider-customer relationships in digital servitization : a relational view on digitalization

(1)

This is a self-archived – parallel published version of this article in the publication archive of the University of Vaasa. It might differ from the original.

Transforming provider-customer relationships in digital servitization : a relational view on digitalization

Author(s): Kamalaldin, Anmar; Linde, Lina; Sjödin, David; Parida, Vinit Title: Transforming provider-customer relationships in digital

servitization : a relational view on digitalization

Year: 2020

Version: Publisher’s PDF

Copyright ©2020 the author(s). Published by Elsevier Inc. This is an open access article under the Attribution–NonCommercial–

NoDerivatives 4.0 International (CC BY–NC–ND 4.0) license, https://creativecommons.org/licenses/by-nc-nd/4.0/.

Please cite the original version:

Kamalaldin, A., Linde, L., Sjödin, D., & Parida, V., (2020).

Transforming provider-customer relationships in digital servitization : a relational view on digitalization. Industrial

marketing management.

https://doi.org/10.1016/j.indmarman.2020.02.004

(2)

Contents lists available atScienceDirect

Industrial Marketing Management

journal homepage:www.elsevier.com/locate/indmarman

Research paper

Transforming provider-customer relationships in digital servitization: A relational view on digitalization

Anmar Kamalaldin

^a

, Lina Linde

^a

, David Sjödin

^a

, Vinit Parida

^a,b,^⁎

aEntrepreneurship and Innovation Group, Department of Business Administration, Technology and Social Sciences, Luleå University of Technology, Luleå 971 87, Sweden

bSchool of Management, University of Vaasa, PO Box 700, FI-65101 Vaasa, Finland

A R T I C L E I N F O Keywords:

Digital servitization Relational view Governance Digital transformation Advanced services Business model innovation

A B S T R A C T

Digitalization is viewed as a source of future competitiveness due to its potential for unlocking new value- creation and revenue-generation opportunities. To profit from digitalization, providers and customers tend to move away from transactional product-centric model to relational service-oriented engagement. This relational transformation is brought about through digital servitization. However, current knowledge about how providers and customers transform their relationship to achieve benefits from digital servitization is lacking. This paper addresses that knowledge gap by applying the relational view theory to a study of four provider-customer relationships engaged in digital servitization. The results provide evidence for four relational components – complementary digitalization capabilities, relation-specific digital assets, digitally enabled knowledge-sharing routines, and partnership governance – that enable providers and customers to profit from digital servitization. A key contribution is the development of a relational transformation framework for digital servitization that provides an overview of how the four relational components evolve as the relationship progresses. In doing so, we contribute to the emerging servitization literature by offering key relational insights into the interdependence of activities throughout the transformation phases of provider-customer relationships in digital servitization.

1. Introduction

We have realized that we need to progress on these issues as we enter digitalization – how do we procure value and how do we work together with our key suppliers? We have thousands of suppliers and all cannot be treated the same – how can we work with our most important partners to increase value creation?

(Chief Procurement Officer of a large mining company) Digitalization is viewed by industry and academia as a source of future competitiveness due to its potential for new value-creation and revenue-generation opportunities. Specifically, digital technologies such as the internet of things, remote monitoring (Grubic, 2014), big data analytics, and artificial intelligence are expected to enable manufacturing companies to undergo the servitization transition from being a product provider to a solutions provider (Hasselblatt, Huikkola, Kohtamäki, & Nickell, 2018;Kohtamäki, Parida, Oghazi, Gebauer, &

Baines, 2019). This rising trend is encapsulated in the concept ofdigital servitization, formally described as the provision of digital services embedded in a physical product (Holmström & Partanen, 2014;

Vendrell-Herrero & Wilson James, 2017). One example is ABB's offering

of a remote optimization service through its collaborative operations centers for gearless mill drives, which allows it to capitalize on its technological expertise by leveraging the efficiencies of digital technology. Typically, product providers adopt a digital servitization strategy to differentiate themselves from competitors (Opresnik &

Taisch, 2015) and to create new revenue streams by establishing closer collaboration with their customers (Scherer, Kloeckner, Ribeiro, Pezzotta, & Pirola, 2016). However, despite the considerable investment in offering digital services, many companies struggle to create real customer value, and both providers and customers ultimately fail to secure a financial return on their investment (Gebauer, Fleisch, &

Friedli, 2005; Kastalli & Van Looy, 2013; Pagoropoulos, Maier, &

McAloone, 2017;Suarez, Cusumano, & Kahl, 2013).

Digitalization combined with servitization significantly transforms provider–customer relationships; and a key challenge for companies pursuing digital servitization is to adapt and revise existing product- centric relationshipsLerch & Gotsch, 2015;Pagoropoulos et al., 2017;

(Sjödin, Parida, Jovanovic, & Visnjic, 2020). This is so because digital services require providers to take on greater responsibility for the core processes of the customer (Lerch & Gotsch, 2015) by shifting from transactional to relational interaction (Reim, Sjödin, & Parida, 2018;

https://doi.org/10.1016/j.indmarman.2020.02.004

Received 21 December 2018; Received in revised form 26 November 2019; Accepted 4 February 2020

⁎Corresponding author at: Department of Business Administration, Technology and Social Sciences, Luleå University of Technology, Luleå 971 87, Sweden.

E-mail addresses:anmar.kamalaldin@ltu.se(A. Kamalaldin),lina.linde@ltu.se(L. Linde),david.sjodin@ltu.se(D. Sjödin),vinit.parida@ltu.se(V. Parida).

Industrial Marketing Management xxx (xxxx) xxx–xxx

(3)

Sousa & da Silveira, 2017). Therefore, digital servitization tends to create closer provider–customer relationships characterized by co- creation logic, long-term commitment, and greater investment in the relationship. Consider a mining equipment provider that goes from selling excavators and haulers to offering site management and optimization of the customer's mining operations. In a case like this, it becomes necessary for both provider and customer to move away from a transactional product-centric model to relational engagement, where the provider works in close collaboration to fulfill the customer's business goals and achieve operational efficiency.

However, the transition to reciprocal long-term relationships is not always easy as, for many companies, this is a step into unknown ter- ritory. For example, companies may struggle with numerous relational issues such as how to balance risk and reward (Reim et al., 2018), how to find the appropriate level of customization, and how to ensure transparency, share data, and integrate digital systems (Coreynen, Matthyssens, & Van Bockhaven, 2017). This underscores the need to investigate and understandhow providers and customers transform their relationships in digital servitization. Accordingly, this study targets two research gaps.

First, prior studies have placed the dominant focus on the provider perspective, with limited insights into the relational aspect (Raddats, Kowalkowski, Benedettini, Burton, & Gebauer, 2019;Sjödin, Parida, &

Wincent, 2016) and how customer organizations must work closely with providers to adapt and, thus, derive benefits from digital technologies (Pagoropoulos et al., 2017). Consequently, this study adopts the theoretical lens of the relational view suggested byDyer and Singh (1998)to better understand how provider and customer relationships facilitate the transformation to digital servitization. Indeed, since this transformation requires engagement and intensive collaboration between provider and customer (Story, Raddats, Burton, Zolkiewski, &

Baines, 2017;Valtakoski, 2017), it is important to understand the requirements of both sides in the partnership. For example, to procure a digital service, such as site management, requires a completely different evaluation process where the value parameter places emphasis on outcome guarantees rather than product features. Consequently, it is necessary to include the less-studied customer perspective to understand the new relational requirements that emerge during digital servitization (Coreynen et al., 2017; Holmlund, Kowalkowski, &

Biggemann, 2016;Tuli, Kohli, & Bharadwaj, 2007;Valtakoski, 2017).

Second, there is a lack of knowledge concerning how provider–customer relationships transform and evolve through digital servitization. Whilst the significance of digital transformation is well established (Ardolino et al., 2018;Porter & Heppelmann, 2015), only a few studies have addressed it from a dynamic relational perspective.

Lerch and Gotsch (2015)have stressed that attempts to offer models dealing with the transformation to advanced digital services have been few. In response, they developed a model that encompasses four generic stages of a company's transformation path from manufacturer to provider of digitalized product–service systems (Lerch & Gotsch, 2014).

However, their model highlights only the provider's transformation journey. A need, therefore, remains to study the transformation of the provider–customer relationship so that the key activities and the dynamics between partners are better understood (Bharadwaj, El Sawy, Pavlou, & Venkatraman, 2013). Arguably, the dynamics in these types of relationship are very different from traditional provider–customer relationships and, therefore, many unforeseen relational challenges can be expected (Reim et al., 2018). However, the change in relational dynamics between providers and customers remains an under-re- searched topic in the servitization literature – a deficiency that is in- creasingly acknowledged (Huang & Chiu, 2018; Reim et al., 2018;

Schuh, Klotzbach, & Gaus, 2008;Sjödin et al., 2016;Sundin, Öhrwall Rönnbäck, & Sakao, 2010).

This study targets these research gaps by applying the relational view (Dyer & Singh, 1998;Dyer, Singh, & Hesterly, 2018) to increase understanding ofhow providers and customers transform their relationships

in digital servitization. It builds on the case studies of four dyadic re- lationships between customers and providers in different industries that have undergone digital servitization. The dyadic cases studied illustrate relational transformation and generate valuable insights into how this transformation is exploited through digital service offerings.

This study contributes to the servitization literature by putting provider–customer relationships in focus and adopting a novel theoretical lens based on the relational view (Dyer, Singh, & Hesterly, 2018;

Dyer & Singh, 1998). This means we also address the call to include and develop relevant theoretical perspectives when studying servitization (Kowalkowski, Gebauer, & Oliva, 2017; Rabetino, Harmsen, Kohtamäki, & Sihvonen, 2018). In addition, we shed light on the emergence of digital servitization, which adds another level of complexity to advanced service provision. Current mainstream servitization literature has just begun to study how firms can benefit from digital servitization. This enables us to add novel knowledge on how provider–customer relationships evolve in the specific context of digital servitization and become transformative, with the explicit goal of creating greater value for both parties.

2. Theoretical background 2.1. Digital servitization

Research has demonstrated a strong interconnection between digitalization and servitization (Gago & Rubalcaba, 2007;Lerch & Gotsch, 2015). The literature points to the fact that digitalization is both a driver and an enabler of servitization (Kohtamäki et al., 2019;Rust &

Huang, 2014;Vendrell-Herrero, Bustinza, Parry, & Georgantzis, 2017).

The usage of digital technologies empowers various types of servitization and service innovation (Gago & Rubalcaba, 2007). Digitalization has actually stimulated companies to move from product-centric models to digital service-oriented offerings (Adrodegari & Saccani, 2017;Ardolino et al., 2018;Rust & Huang, 2014).Iansiti and Lakhani (2014)argue that digital transformation changes the customer's value proposition; it alters how a company creates and captures value since digitalization principally involves the provision of services. The opportunities to expand services increase when companies synchronize digitalization, connectivity, and data analytics (Martín-Peña, Díaz- Garrido, & Sánchez-López, 2018).

A research sub-stream in the servitization literature has been named

‘digital servitization’. It can be defined as “the transformation in processes, capabilities, and offerings within industrial firms and their associated ecosystems to progressively create, deliver, and capture increased service value arising from a broad range of enabling digital technologies such as the Internet of Things (IoT), big data, artificial intelligence (AI), and cloud computing” (Sjödin, Parida, Kohtamäki, &

Wincent, 2020). Digital servitization involves the utilization of digital tools for transforming a product-centric business model to a service- centric logic ((Sklyar et al., 2019)). This owes to the evolution of ‘smart, connected products’ – a combination of hardware, software, sensors, data storage, and connectivity – which have transformed manufacturing companies (Porter & Heppelmann, 2014; Porter &

Heppelmann, 2015). An example of digital services is remote monitoring, which is considered a key enabler of servitization since it is vital to remotely monitor the product's location, condition and use (Baines &

Lightfoot, 2013;Ulaga & Reinartz, 2011). Through remote monitoring and diagnostics, the company can preemptively repair a machine prior to failure, rather than reactively mend it after it has shut down (Allmendinger & Lombreglia, 2005). Through connecting digital and physical systems, remote services create the opportunity to provide availability guarantees, for example (Lerch & Gotsch, 2015). Never- theless, digital servitization creates both opportunities and challenges for companies thus engaged.

It is worth noting that digital opportunities arise at a speed that many companies are unable to cope with expeditiously. Consequently,

(4)

companies need to embark on an efficient digital transformation that allows them to take full advantage of digitalization and servitization.

However, many companies struggle with digital transformation because it requires a change in provider–customer relationships whereby new and innovative approaches need to be adopted (Iansiti & Lakhani, 2014). These relationships should be looked at in a dynamic rather than a static way in order to maximize value for both parties. However, there are few insights in the literature on how provider–customer relationships should be transformed in the context of digital servitization.

Iansiti and Lakhani (2014)note that a transformation journey can go through three phases in the search for increased value. First, it starts as a transactional relationship between provider and customer. Second, it evolves into a contractual relationship, where the risk is shared and the total cost of ownership is reduced. Third, the relationship develops further with expanded customer outcomes as assets and operations are optimized using data and analytics. However, knowledge is lacking on how to achieve the objectives of each phase, and how to move from one phase to the next. In other words, we need a better understanding of what aspects ought to be considered throughout this transformation in order to attain high relational value. This provides the primary moti- vation for studying digital servitization from a relational view.

2.2. A relational view on digital servitization

This study applies the theoretical perspective of the relational view (Dyer, Singh, & Hesterly, 2018; Dyer & Singh, 1998) to the study of digital servitization relationships. A relational viewpoint argues that competitive advantage is a result of mutually adapted inter-firm rela- tions and the joint input of partners (Dyer & Singh, 1998;Lavie, 2006), which enables companies to co-evolve so that relational rents are generated. Although competition between companies might still be the general rule, firms that integrate their resources in a distinctive way may secure greater advantage compared to competing firms unable or unwilling to do so (Dyer & Singh, 1998).

The importance of a relational view is particularly relevant in the context of digital servitization (Cenamor, Sjödin, & Parida, 2017;

Eloranta & Turunen, 2015) because the implementation of integrated products and services can only succeed when both provider and customer deploy them, and not simply because a provider delivers them (Tuli et al., 2007). Thus, providing this kind of service can be seen as a

‘longitudinal relational process’ between provider and customer (Storbacka, Windahl, Nenonen, & Salonen, 2013). Digital services are significantly transforming inter-firm relationships and influencing governance patterns across companies (Bharadwaj et al., 2013). Pro- viders wish to increase customer value, and so they integrate themselves into their customers' business processes (Matthyssens &

Vandenbempt, 2008). A major issue in a servitization transformation is the quality of interaction between provider and customer so that customized and comprehensive solutions that offer real value are provided (Kohtamäki, Partanen, Parida, & Wincent, 2013;Lerch & Gotsch, 2015;

Reim, Parida, & Örtqvist, 2015; Viljakainen & Toivonen, 2014). The closeness and quality of this interaction is key to value co-creation Grönroos & Voima, 2013; (Sjödin, 2019) because the development of customized solutions needs a collaborative innovation approach as well as connected production resources shared between partners (Martín- Peña et al., 2018).

This study follows Dyer and Singh (1998) who suggest four determinants of inter-organizational competitive advantage: com- plementary resources and capabilities, relation-specific assets, knowledge- sharing routines, and effective governance. We argue that these determi- nants hold significant explanatory potential for understanding how inter-firm relationships are transformed through digital servitization.

Indeed, as Dyer and Singh (1998) suggest, these determinants can generate relational rents or the “supernormal profit jointly generated in an exchange relationship that cannot be generated by either company in isolation and can only be created through the joint idiosyncratic

contributions of the specific alliance partners” (Dyer & Singh, 1998, p.

662). However, these relational rents will not remain static over time, especially in times of industrial disruption such as the current digitalization of industry. Thus, it is important that partners dynamically consider these determinants in order to create and capture value over time (Dyer, Singh, & Hesterly, 2018), and to be able to capitalize on emerging digitalization opportunities (Sjödin, Parida, Leksell, &

Petrovic, 2018). In the paragraphs that follow, we describe these four determinants of the relational view and how they may be con- ceptualized in the context of digital servitization.

First, Dyer, Singh, & Hesterly, 2018 argue that access to com- plementary resources and capabilitiesis considered the initial rationale in forming a partnership. In this case, the marginal return on a partner's resources increases in the presence of resources from the other partner (Hess & Rothaermel, 2011;Milgrom & Roberts, 1995). Dyer and Singh argue that “the greater the proportion is of synergy-sensitive resources owned by alliance partners that, when combined, increase the degree to which the resources are valuable, rare, and difficult to imitate, the greater the potential will be to generate relational rents” (1998, p. 667).

Partnerships allow companies to acquire assets, competences or capabilities – in particular, specialized expertise (Oliver, 1997). Having competences, experience and knowledge are essential in implementing digital technologies (Ardolino et al., 2018;Cenamor et al., 2017). So, when companies do not have the necessary capabilities or resources, they tend to fill the gap by partnering with other companies. Research has shown that value is co-created by providers and customers through integrating their resources and exploiting their shared competences (Grönroos, 2011; Grönroos & Voima, 2013;Vargo, Maglio, & Akaka, 2008). Customers typically seek to involve providers in operations outside their own core competences (Sjödin et al., 2018). However, research on digitalization capabilities has concentrated on the provider's perspective e.g. (Sjödin, Parida, & Kohtamäki, 2016), with only limited consideration given to the customer's viewpoint.Lenka, Parida, and Wincent (2017), for instance, highlight the digitalization capabilities that providers must develop in order to interact and co-create value with their customers; intelligence capability, connect capability, and analytic capability. However, there is limited research on how a company's capabilities complement its partners' capacities and resources within digital servitization (Pagoropoulos et al., 2017). Again, the importance of looking at this aspect from the perspective of a relational view is stressed.

Second, Dyer and Singh argue that “the greater the alliance partners' investment is inrelation-specific assets, the greater the potential will be for relational rents” (1998, p. 664). These assets are usually specialized and are considered to be of strategic importance for the relationship – necessary conditions for generating relational rent (Amit &

Schoemaker, 1993). Williamson (1985) names three types of asset specificity. First, site specificity, which refers to locating successive production stages close to one another. Second, physical asset specificity, which refers to transaction-specific or tailored capital investments such as customized machinery. Third, human asset specificity, which refers to transaction-specific know-how that can be garnered from long- term relationships and staff dedicated to those relationships. Provision of advanced services naturally involves investing in relation-specific investments and co-specialized assets. For instance,Sjödin et al. (2016) discuss the importance of building knowledge about partner operations and roles to clarify and redefine relationships. In terms of digital servitization, the more digital intensity increases within a company's business strategy, the more likely its scaling options will be based on a partnership with other companies by means of shared digital assets (Bharadwaj et al., 2013). Offering availability guarantees for machines and plants, for example, requires linking customer plants to the provider's digital architecture through a compatible communication network (Lerch & Gotsch, 2015). Yet, how relation-specific digital assets evolve throughout the relationship needs to be investigated by further research.

(5)

Third, Dyer and Singh emphasize that “the greater the alliance partners' investment is in inter-firm knowledge-sharing routines, the greater the potential will be for relational rents” (1998, p. 665). They define knowledge-sharing routines as a regular pattern of interactions between companies that allow specialized knowledge to be transferred, recombined, or created (Grant, 1996). In other words, they are pur- posefully designed processes to facilitate knowledge exchange between partners. In the digital era, technologies allow companies to easily communicate and share in-depth information and knowledge through digital means (Gago & Rubalcaba, 2007; Martín-Peña et al., 2018).

However, it is not given that this translates into enhanced knowledge sharing or performance. Information overload is a common term for paralysis arising from too much data that is hard to prioritize and de- code. These problems are increasing in digitalization where the cost of additional data collection is next to none. Thus, the use of data is a key aspect to consider as digitalization lays the foundation for translating digital data into knowledge (Barney, Wright, & Ketchen Jr, 2001) and leads to improved transparency and better-informed decision making (Ness, Swift, Ranasinghe, Xing, & Soebarto, 2015). Thus, digitalization can enable innovation by facilitating knowledge exchange and ab- sorptive capacity between actors (Barrett, Davidson, Prabhu, & Vargo, 2015;Carlo, Lyytinen, & Rose, 2012;Joshi, Chi, Datta, & Han, 2010;

Slaughter & Kirsch, 2006). It is obviously not feasible to have em- ployees aggregate and analyze the real-time data generated from smart machinery, such as the physical location of a product or the tempera- ture of a component. Instead, this is enabled by digital services that depend on ‘machine intelligence’ where data points are automatically gathered, validated, stored, and turned into information that can be acted on (Allmendinger & Lombreglia, 2005). What really matters is transforming this data into valuable insights and actions (Lenka et al., 2017). However, further research is needed to understand how partners must work together in this endeavor.

Fourth, in recent developments on the relational view (Dyer, Singh,

& Hesterly, 2018), theeffective governanceof relationships is regarded as the key differentiator that allows the development of the other determinants. Governance can be considered the safeguard used by partners to enforce what they have agreed, and it is intimately connected to all other determinants of relational rents that partners must govern as part of their relationship (Dyer, Singh, & Hesterly, 2018;Dyer

& Singh, 1998). Dyer and Singh (1998) differentiate between two classes of governance. The first is governance that depends on ‘third- party enforcement’ of agreements, such as legal contracts where the state, for example, can be the third-party enforcer in the event of dis- pute. The second class of governance depends on ‘self-enforcing’

agreements, where no third party intervenes. Self-enforcement can be either through ‘formal’ safeguards such as financial penalties (Reim et al., 2018; Williamson, 1983), or by ‘informal’ safeguards such as goodwill, trust (Gulati, 1995;Powell, 1990;Uzzi, 1997) and reputation (Larson, 1992; Weigelt & Camerer, 1988). Generally, self-enforcing mechanisms are more effective than third-party enforcement (Dyer &

Singh, 1998) because they enable greater flexibility and innovation in the relationship (Kohtamäki et al., 2013; Reim et al., 2018). Yet, partnerships mostly start with formal mechanisms and then gradually adopt more informal means as the relationship develops (Gulati, 1995).

Dyer and Singh suggest that “the greater the alliance partners' ability to align transactions with governance structures in a discriminating (transaction cost minimizing and value maximizing) way, the greater the potential will be for relational rents” (1998, p. 669). In the context of digital servitization,Svahn, Mathiassen, and Lindgren (2017)argue that a key paradox experienced by companies relates to governance – that is, achieving balance between control versus flexibility during the transformation process. As digital servitization is largely based on innovation, there is a major need to balance new opportunities and established practices, and governance mechanisms should enable crea- tivity and exploration of digital opportunities (Svahn et al., 2017).

Sarker, Sarker, Sahaym, and Bjørn-Andersen (2012) also stress that

governance mechanisms are considered one of the enablers and in- hibitors of value co-creation between partners, especially in the information technology context that they studied. The servitization literature lays particular stress on the importance of self-reinforcing mechanisms, such as mutual trust and the extent of partner commitment, since these reduce bureaucratic complexity and, in consequence, transaction costs (Reim et al., 2018;Sjödin et al., 2016). However, the authors do not disregard the role of contractual agreements since they help to combat opportunism by explicitly stating terms, conditions, and responsibilities in the partnership – hence, leading to more effective value co-creation (Park & Ungson, 2001;Poppo & Zenger, 2002;Reuer

& Arino, 2007;Sjödin et al., 2016). Nevertheless, digital servitization contracts should not replicate traditional cost-control contracts. In- itiatives should be taken to develop new generic contracts that emphasize mutual liability and incentives for co-creating digital services with partners (Svahn et al., 2017), while maintaining acceptable control over value appropriation (Boudreau, 2010). Although the literature underscores the importance of maintaining a balance between control and flexibility in governing digital servitization partnerships, there is scant knowledge on how this balance should be fostered as the relationship matures over time.

In short, we consider the relational view to be a very useful theoretical lens to study how a provider–customer relationship is transformed in the context of digital servitization. The relational view offers a better, and more dynamic, theoretical lens compared to the resource- based view (e.g.Dyer, Singh, & Hesterly, 2018). Although the resource- based view highlights how a company maintains its competitive advantage by obtaining valuable, rare, inimitable, and non-substitutable resources (Barney, 1991), it does not address the fact that these resources can extend beyond the boundaries of a single company and can be combined with a partner's resources. Given the speed of development in the digital era, it is clear that no company can keep pace on its own (Bogers, Chesbrough, & Moedas, 2018). Hence, the relationship between providers and customers is an important unit of analysis in examining value creation and profit maximization (Dyer & Singh, 1998) in digital servitization. As the partnership transforms over time, this study aims to research how each of the four determinants of relational rent evolves throughout the relationship.

3. Research methods

3.1. Research approach and case selection

This study is based on an exploratory multiple case study (Yin, 2009) of four business-to-business (B2B) dyadic relationships between providers and customers of digital services. The study seeks to shed light on how their relationships transform over time in the context of digital servitization. Hence, the unit of analysis is the dyadic relationship between two companies. We decided to apply a multiple case study approach as it is commonly used in the industrial marketing research domain (Halinen & Törnroos, 2005). This case study approach makes it possible to mobilize multiple observations on complex relational processes (Eisenhardt, 1989; Eisenhardt & Graebner, 2007), which are particularly useful in developing new insights into theoretically novel phenomena (Edmondson & McManus, 2007), such as how providers and customers transform their relationships in digital servitization.

Thus, the possibility of obtaining versatile and complementary insights is opened up.

Building on recommendations suggested by Glaser and Strauss (1967), we opted for theoretical sampling in order to select cases that would illuminate how companies transform their relationships when undergoing digital servitization (Suddaby, 2006). The case selection criteria were informed by the study's research question: how do providers and customers transform their relationship in digital servitization? The question highlights three aspects: the context of digital servitization, the transformation of the relationship, and the dyadic

(6)

provider–customer relationship. Accordingly, the case selection criteria are as follows.

First, we selected globally active Swedish providers and customers engaged in longstanding relationships in digital servitization. The selected companies represent diverse industries – namely manufacturing, telecom, forestry, energy, and mining industries – which provides the opportunity to contrast various industrial perspectives on relational processes.

Second, these relationships have been developing over time with notable progress. All cases are similar in the sense that they underwent a transformation from product-centric models to digital service-oriented offerings, thus facilitating the study of how they evolve and transform. So, a key selection criterion was the ability of the provider and customer to vividly describe the relationship trajectory and provide in-depth information about the relationship and its key activities, supported by documents and background information.

Third, we selected cases where we had established good contacts with both provider and customer in the relationship due to the ongoing nature of the research project, as this made for rich data collection.

Unlike other studies that ignore the customer perspective, we followed the example of Tuli et al. (2007)in collecting dyadic data (i.e. from both customer and provider views) on the evolution of the relationship, enabling us to gain a deeper understanding of the interactive relationships relevant to the digital servitization context.

3.2. Data collection

Data for the present study was gathered primarily through in- dividual, in-depth interviews with participants in the four relationships.

We developed a semi-structured interview tool for our interviews. The unit of analysis was the relationship between provider and customer.

Therefore, we undertook interviews with numerous managers from both the customer and the provider sides of the relationship. To do so, we organized separate interviews with each informant.

In total, 40 informants were interviewed from all cases. We began by interviewing key informants – senior executives from case companies – who were actively involved in the relationship. Additional informants were identified by using a snowballing technique where key informants were asked to recommend people who had an active role in the relationship and were able to describe how the relationship had progressed. To capture a multifaceted view, we interviewed individuals exercising various functional roles for providers and customers engaged in the relationships. This was deemed necessary since digital servitization relationships typically require complex interactions between multiple organizational functions. The informants interviewed included business developers, R&D managers, project managers, production managers, product managers, plus maintenance and technical support staff. This allowed us to obtain a wider understanding of the cases from different perspectives.Table 1gives an overview of the cases and the positions of company informants interviewed.

Informants were asked open-ended questions with the support of an interview guide (example of interview questions can be found in Appendix 1). The guide was based on themes about digital servitization, value co-creation between provider and customer, and how business relationships start and evolve over time. For example, informants were asked to consider questions relating to broad themes such as ‘How did the digital servitization relationships evolve?’, ‘Which activities are critical in facilitating digital servitization?’, and ‘How was the relationship governed?’. In seeking answers to these overarching questions, we encouraged informants to base their answers not only on the relationships studied but also on their broader experience so that empirical comparisons were facilitated. Follow-up questions were asked for clarification and to obtain further details, which allowed further exploration of interesting relevant cases. The interview guide was also revised continuously as we derived new insights from the interviews

and secondary data – thus, increasing relevance and deepening Table1 Overviewofstudiedrelationships,companiesandinformants. CaseRelationshipdescriptionCompanypseudonymIndustryNo.ofemployeesInformants R1JointefforttoconnectandintegrateBeta'sfleetofAlpha's machinesandcontrolsystemstoidentifyoptimization opportunitiesandfacilitateincreasedservicelevel

Alpha(Provider)Power&automationtechnology7800Total:6(2accountmanagers;2businessdevelopmentmanagers;1headofservice center;1service&supportmanager) Beta(Customer)Mining5700Total:9(1chiefprocurementofficer;1headofprocurementdevelopment;3 procurementmanagers;1R&Dmanager;1automationmanager;1technology developmentengineer;1maintenancemanager) R2Jointeffortstoconnectandintegrateinformationfrom Gamma'soverallproductionanddistributionnetworkto visualizesystemperformanceandidentifyimprovements Alpha(Provider)Power&automationtechnology7800Total:2(1digitalizationmanager;1productmanager) Gamma(Customer)Energy&utilities700Total:3(2ITprojectleaders;1strategymanager) R3Jointdevelopmentofadigitalserviceplatformanda predictivemaintenanceserviceoffertousedataand analyticstolowerlifecyclecosts

Delta(Provider)Machinerymanufacturing600Total:8(1marketingmanager;3salesmanagers;2servicebusinessplanning managers;2technicalmanagers) Epsilon(Customer)Forestry4100Total:6(1chieftechnicalofficer;2technicalmanagers;2businessdevelopment managers;1productionmanager) R4Partnershipdesignedtoimplementthelatesttechnology anddigitalsolutionsandcontinuouslyimprovenetwork operations Zeta(Provider)Telecomequipment12,700Total:4(1businessmodelsresearcher;1headofstrategy&businessdevelopment;1 commercialmanagementdirector;1businessoperationsmanager) Eta(Customer)Telecom17,700Total:2(1businessdevelopmentmanager;1contractmanager)

(7)

understanding. Interviews took approximately 60–120 min each. All interviews were recorded and transcribed, and transcripts provided the basis for data analysis.

We triangulated our data by applying multiple data-collection techniques, including multiple interviews and a review of documents (Jick, 1979). Multiple data sources were also leveraged to distinguish phases of the relationship; data represented multiple periods of time although they were all collected contemporaneously. We reviewed company reports, agreements, and project documents to validate and provide context to our informants' views, thus enabling empirical tri- angulation. For example, an updated documented process of Beta on how to interact with suppliers when aiming to develop digital service offerings validated and helped to refine initial codes that had been derived from interviews. This was also studied in comparison to pre- vious practices of interaction with suppliers.

By using multiple sources of evidence – i.e. both different informants and different secondary sources – we were able to increase construct validity since these insights gave a nuanced picture of the phenomenon studied and the conclusions to be drawn (Yin, 2009). The initial results of the study were presented to a number of informants from case companies in order to increase validity. Only minor revisions were introduced during these interactions. Furthermore, to increase reliability and enhance transparency, as well as the possibility of re- plication, a case-study protocol with emphasis on field procedures and case-study questions was constructed along with a case-study database.

The aim was to keep track of the process and allow multiple researchers

to collect and analyze data. The database included physical and digital material such as case-study notes, documents, and analysis.

3.3. Data analysis

The data analysis was based on a thematic analysis approach, which provides ways to identify patterns in a large and complex dataset (Braun & Clarke, 2006). Moreover, it provides a means to effectively and accurately identify links within analytical themes. Through a series of iterations and comparisons, it is possible to identify themes and overarching dimensions so that an empirically grounded framework can be developed. In doing so, we followed a three-step process similar to that described in recent literature. Data was coded into categories following a thematic analysis approach; these were then clustered into second-order themes, which were converged into aggregate dimensions (Braun & Clarke, 2006;Gioia, Corley, & Hamilton, 2013). A more de- tailed description of these steps is included below.

The first step in our data analysis focused on an in-depth analysis of raw data (e.g. interview transcripts). This analysis focused on reading every interview several times, and marking phrases and passages re- lated to the overarching research purpose. By coding the common words, phrases, terms, and labels mentioned by informants, it was possible to identify first-order categories of codes, which express the views of the informants in their own words. This was facilitated by MAXQDA software and resulted in first-order categories.

The second step of the analysis built on the first-order categories Fig. 1.Data structure and coding process.

(8)

Table2 Overviewoftherelationaltransformationinstudieddigitalservitizationrelationships. CaseComplementarydigitalizationcapabilitiesRelation-specificdigitalassetsDigitallyenabledknowledge-sharingroutinesPartnershipgovernance R1▪Beta'sknowledgeonminingprocesseswas complementedwithAlpha'sknowledgeon digitalminingequipmentandcontrolsystems. ▪Complementaritywasreassessedforeachnew project.

FOUNDATIONALPHASE Alpha(provider) Beta (customer) Digitalinfrastructurewasbuiltandconnectedwith Alpha'sequipmentandsystems,whichwere managedbydedicatedengineers.

DatawascollectedfromequipmentthroughIoT. Discussionsandfeedbackamongoperatorsand engineerstookplacetoexplorehowtoactuponit.

Responsibilitiesandrightsweremeticulously formalizedinthecontractwithhighcontrollevel. INTERMEDIATEPHASE Alphatailoredadigitalplatformthatconnects smoothlytoBeta'ssystems,andfurtherresources wereallocatedtoreexamineon-siteprocesses.

Datafromallmachineswasconnectedand accumulatedinaninterfaceforbenchmarkingtheir performanceandimpactoneachother.Daily operationalmeetingsandmonthly/quarterly managerialmeetingswereset.

Revisionofthecontractandexperimentswith differentcontractualmodelsbasedon‘gain/pain sharing’logic. ADVANCEDPHASE Digitalplatformwasfurtherdevelopedtoenable customizedoperationalsolutionsassessedbyateam ofdatascientists.

AlphaandBetaexchangedinformationforfurther fleetoptimizationtothebenefitofboth.Ajoint analysisteamsuggestedimprovementopportunitiesto management.

Performance-basedcontractandajoint developmentstrategyformoreadvanceddigital services. R2▪Gamma'sknowledgeongridmanagementwas complementedwithAlpha'sdigitalexpertisefor developingthesystem. ▪Complementaritywasmonitoredthroughpilot projects.

FOUNDATIONALPHASE Alpha (provider) Gamma (customer) Alpha'sapplicationswerebuiltonGamma'sdigital systems,anddesignatedengineerswereassignedfor collaborativeoperations.

Databasewasestablishedforcollectingdatafrom machinestotracktheiractivityandmonitorthe energysystem'sstatus.Ad-hocdiscussionswereheld forimprovingprocesses.

Thedetailsofthecollaborationmodelweredefined inthecontract. INTERMEDIATEPHASE DigitalplatformwasdevelopedbasedonGamma's needs,andacooperativedigitalizationcenterwas established.

Datawasaccumulatedfromallmachinesfor condition-basedmaintenanceandoptimization. Managerialdiscussionsofjointoperationswereheld semi-annually.

Contractrevisedtoredefineperformanceindicators andrealigngoalsandincentives. ADVANCEDPHASE Digitalplatformwasfurtherdevelopedtoenable identifyingoperationalproblemssuchaspositionsof waterleakages.Ajointteamwasestablishedfor developingsolutions.

ThedataoftheAlpha'swiderfleetofmachines supplementedGamma'sdatatoidentifyfurther improvementopportunities,whichwereevaluated andimplementedbyjointteams.

Jointgovernancethroughasteeringgroupthatsets long-termstrategicdirectionsandajointbudget. R3▪Epsilon'sknowledgeoncostefficiencyoflogging wascomplementedwithDelta'stechnical expertiseonefficientuseofsmartmachinery. ▪Complementaritywasreassessedforensuring machines'andoperators'efficiency.

FOUNDATIONALPHASE Delta (provider) Epsilon (customer) Allmachineswereequippedwithsensorstoconnect thewholemachineryfleet,anddedicatedmachine instructorswereassigned.

Theinstalledsoftwareandhardwareenabledthe monitoringofmachines'performanceandcalibration. Ad-hocdiscussionsofproductionefficiency.

Contractwassetupbasedonproductsalesandafter salesservicepackage.Goalsandfollow-up mechanismswerespecifiedinthecontract. INTERMEDIATEPHASE Allmachineswereconnectedtoonedigitalplatform forhandlingoperations,andstaffwereassignedfor assessingoperations'efficiency.

Datawasaccumulatedfromdiversemachinesof differentbrandsforfacilitatingbettersite management.Performanceimprovementwas discussedbetweenEpsilon'soperatorsandDelta's mechanicssemi-annually.

Productrelationshipwastransformedtoaservice contract,enablingproactivemaintenanceand improvedmachineperformance.Priceswaere revisedaccordingly. ADVANCEDPHASE Machineoptimizationserviceswerelaunchedbased oninsightsfromthedigitalplatformthatwere analyzedbyajointoperationalcontrolteam.

Delta'sdatawasintegratedwithEpsilon'sdataand usedforoperatortrainingprogramanddigitalservice package.Ajointteamregularlydiscussed opportunitiesforfurtherdataintegrationandanalyzed latestinnovationsinforestryindustry.

Re-evaluationoftherelationshiptobuildmutual understandingandtrust,withafocusonoutcomes ratherthanthecontractinitself. (continuedonnextpage)

(9)

and used further analysis to discover links and patterns within them.

This iterative process led to the formation of second-order themes that represent theoretically distinct concepts created by combining first- order categories. Our analysis identified 6 second-order themes, which were at a higher level of abstraction compared to the first-order categories. These themes relate to various approaches that enable the relationships to progress. In accordance with validity claims in the literature, the themes were further refined, based on insights from prior literature as well as data from interviews and secondary sources such as internal documents, presentations, newspapers, and company websites (Kumar, Stern, & Anderson, 1993). This step of the data analysis was conducted conjointly by the authors, who thoroughly discussed the data structure.

The third step involved the generation of aggregate dimensions that represented a still-higher level of abstraction in the coding. Similar to other studies (e.g.Einola, Kohtamäki, Parida, & Wincent, 2017;Lenka, Parida, Sjödin, & Wincent, 2018;Reim, Sjödin, & Parida, 2019) that followed the approach ofGioia et al. (2013), we used insights from the literature to guide the formation of theoretically rooted dimensions. In the analysis of the themes, we evaluated different theoretical frame- works but found that the data were closely aligned with the theoretical framework of the relational view (Dyer & Singh, 1998), which then provided the structure for how the themes converged into dimensions.

Consequently, we identified the following aggregate dimensions: complementary digitalization capabilities, relation-specific digital assets, digitally enabled knowledge-sharing routines, and partnership governance. Thus, the aggregate dimensions represent a theoretically and empirically grounded categorization. Fig. 1 shows the entire data structure resulting from the data analysis.

As a fourth and final step, we assessed the progression across the studied relationships as we sought to uncover how digital servitization relationships unfold and how companies transform them. During this formative step, each researcher independently categorized the first- order categories and its associated data for each case with the relationship's progress category (foundational phase, intermediate phase, advanced phase). The researchers then came together to compare the results of categorizing the data that contributed to building each phase.

Generally speaking, there was considerable agreement between authors on the independently categorized data. In cases of disagreement, we discussed the data and the reasoning behind the choice of phase category, which led to an agreement on establishing the connection between first-order categories and different phases of the relationships.

Table 2provides a summarized overview (i.e. cross-case analysis) of the progression of provider-customer relationship within each case.

Although the identified patterns were largely shared across the studied relationships, some were more evident within specific cases, and some took different form or focus among cases. The specifics of each case can be apprehended through a horizontal view ofTable 2, whilst a vertical view of the table can enable a comparison of the cases across the aggregate dimensions. This is supported by empirical evidence showing representative quotations from each case for each of the first-order categories (seeAppendix 2). This practice of comparing cases allowed us to further refine our data structure and generate an overall framework (Nag, Corley, & Gioia, 2007), explaining how the relationships unfold by linking various phenomena that emerged from the data analysis. Hence, a theoretically and empirically grounded framework was developed (seeFig. 2) through theorizing the logic and linkages across aggregate dimensions, second-order themes, and first-order categories.

4. Findings

Several insights emerged from studying the evolution of digital servitization relationships between provider and customer. Table 2 provides a simplified overview of the analysis, which shows that four aggregate dimensions make up the core relational components Table2(continued) CaseComplementarydigitalizationcapabilitiesRelation-specificdigitalassetsDigitallyenabledknowledge-sharingroutinesPartnershipgovernance R4 Zeta (provider) Eta (customer)

▪Eta'sbusinessknowledgewascomplemented withZeta'sknowledgeonefficientuseoftelecom equipment. ▪Complementaritywasmonitoredthrough regularfollowupmeetings.

FOUNDATIONALPHASE Investmentininstallingbasestationsanddigital systems.TheZetaassigneddedicatedstafffor establishingprocessesandserviceintervals.

Datawascollectedtounderstandhowtorunnetwork operationsefficiently.Discussionsoncapacity improvementswereundertakeninunstructured manner.

Boundaryconditionsandback-stopswereidentified andincorporatedintothecontract. INTERMEDIATEPHASE Digitalplatformwasdevelopedforconnectingall operationalsystems.Resourceswereallocatedto developdigitalandbusinesscapabilities.

Datafromalloperationalsystemswereaccumulated andconnectedtoderiveinsightsforoptimizing operations,whichwerediscussedandanalyzedin formalizedregularinteractions.

A‘reward-penalty’logicwasincorporatedintothe contracttoalignincentives. ADVANCEDPHASE Thedigitalplatformbecameacentralpartfor improvingnetworkoperationsandZetawasgiven freedomtodesigncustomizeddigitalsolutions whichwerereviewedbyajointcollaborativecenter.

Furtherdatatransparencywasencouragedbyopen discussionontheexpectationsandneedsofeach party.Ajointteamofstafffromdiversefunctions workedtogetherforcontinuousevaluationand improvement.

Emphasisonrelationalbenefitsandmaintaininga win-winincontractimplementation.